LoveQuant

Don't guess the relationship. Read the signal.
别凭感觉，用数据读懂关系。

LoveQuant is an open-source relationship signal board that turns messaging rhythm into candlesticks, overlap heatmaps, delay distributions, anomaly markers, and action-ready reports.

This repository currently ships the interactive concept site, visual system, and product narrative for LoveQuant: a self-hosted, cross-channel relationship analytics experience designed for WeChat, WhatsApp, Telegram, iMessage, LINE, Messenger, Instagram DM, and structured imports.

Live Demo · Market Poster · Teaser GIF · Design Philosophy · Algorithmic Philosophy

Why LoveQuant

• Read relationship momentum with data instead of guesswork
• See timing, message investment, initiative balance, overlap windows, and anomaly shifts in one signal language
• Support the channels couples actually use, from WeChat and WhatsApp to Telegram, LINE, iMessage, and manual exports
• Keep the stack self-hosted, inspectable, and extensible

Highlights

• Market-style K-line timeline for 28-day relationship heat
• 7-day overlap heatmap with visible hot zones and cooling windows
• Reply delay distribution and message length trend analysis
• Initiative balance and health scoring on a single dashboard surface
• Parser demo that shows raw messages, extracted signals, and runtime flow together
• Open-source runtime map for connectors, scoring, reports, and future extensions

Core Signals

Signal	What it reveals	Surface
Message Frequency	Relationship activity trend	Candlestick / trend
Reply Delay	Responsiveness and priority shifts	Distribution
Message Length	Conversation investment level	Trend
Initiative Balance	Who is carrying the rhythm	Split bar
Active Time Overlap	Shared availability windows	Heatmap
Health Score	Overall relationship state	Gauge
Anomaly Alerts	Sudden drops, spikes, or collapses	Marker list
AI Suggestions	What to do next	Action prompts

Channels and Input Modes

Channels

• WeChat
• WhatsApp
• Telegram
• iMessage
• Messenger
• LINE
• Instagram DM
• Exported history / manual summaries

Input Modes

1. Shared thread sync Metadata from a shared conversation or workspace flows into the dashboard.
2. Manual import / export Summaries, exports, or snapshots update the report immediately.
3. Presence + time-window signals Lightweight overlap signals for channels where full sync is not ideal.

What This Repo Contains Today

• A polished interactive homepage in lovequant-demo/
• A GitHub Pages deployment workflow in .github/workflows/deploy-pages.yml
• Visual and narrative references in themes/, art/, and assets/
• Product positioning, brand direction, and signal design notes in this root directory

Self-Hosted Model

LoveQuant is designed to be run on your own stack.

• Bring your own adapters for exports, APIs, bots, or manual imports
• Keep ingestion, normalization, scoring, and reporting logic auditable
• Control storage format, retention, and report outputs yourself
• Extend the dashboard into reports, alerts, bots, or custom views

Repo Layout

.
├── .github/workflows/      # GitHub Pages deployment
├── art/                    # Concept artifacts and visual references
├── assets/                 # Poster and teaser assets
├── lovequant-demo/         # Vite + React interactive site
├── themes/                 # Theme notes and direction
├── DESIGN_PHILOSOPHY.md    # Visual system notes
└── ALGORITHMIC_PHILOSOPHY.md

Development Direction

The current repo focuses on the product surface first: narrative, interaction, dashboard language, and the self-hosted mental model.

The next layer is runtime implementation:

• source adapters
• metadata ingestion
• scoring engine
• report generation
• export and automation hooks

License

This project is licensed under the MIT License.